Bonner Zoologische Beiträge

ZOBODAT - www.zobodat.at

Zoologisch-Botanische Datenbank/Zoological-Botanical Database

Digitale Literatur/Digital Literature

Zeitschrift/Journal: Bonn zoological Bulletin - früher Bonner Zoologische Beiträge.

Jahr/Year: 1996

Band/Volume: 46

Autor(en)/Author(s): Palacios Fernando

Artikel/Article: Systematics of the indigenous hares of Italy traditionally identified as Lepus europaeus Pallas, 1778 (Mammalia: Leporidae) 59- 91 © Biodiversity Heritage Library, http://www.biodiversitylibrary.org/; www.zoologicalbulletin.de; www.biologiezentrum.at

Bonn. zool. Beitr. Bd. 46 H. 1-4 S. 59-91 Bonn, Juni 1996

Systematics of the indigenous hares of Italy traditionally identified as Lepus europaeus Pallas, 1778 (Mammalia: Leporidae)

Fernando Palacios

Abstract. Hares living on the Italian Peninsula and in Sicily have traditionally been identified as a single species (Lepus europaeus Pallas, 1778) but a taxonomic study based on the size, morphology, and pelage of indigenous specimens of this area held in museum collections clearly reveals that they belong to two different species, L. europaeus and Lepus corsicanus de Winton, 1898. The geographic localities of the specimens indicate that they occupied parapatric areas in natural conditions, L. europaeus in the northern part of the Italian Peninsula and L. corsicanus in the southern part, being separated between Siena

and Rome. The two species are very distinct phenetically and the species differentiation is

possible using both univariate and multivariate statistics. There is a notable uniformity in the phenetic characteristics of each species within the natural areas of distribution and no evidence of hybridization has been found as the specimens do not have intermediate characters even in neighboring localities within their geographic areas. Among the discriminant features determining the distinctiveness of these two species the color of the

basal fringe of the dorsal underfur of adult specimens which is always white in L.

europaeus and grey in L. corsicanus is worthy of note. As for skull size, there are several measurements which do not overlap, L. corsicanus being smaller than L. europaeus. As regards dental characters, the difference in the shape of the posterior contour of the cross

section of the first upper incisor, which is concave in L. europaeus and smooth or convex

in L. corsicanus, is important. Statistical comparisons of frequencies of skull and dental character states showed high Chisquare values and very low probabilities suggesting that

there is a significant difference between the two species. Two correspondence factor analyses based on the above frequencies served to separate well the specimens of the two species into different groups without plot overlapping. As regards phenetic relationships of L. corsicanus with other SW European hares, L. castroviejoi from the Cantabrian Mts showed the most extensive phenetic similarity in a discriminant analysis made with skull measurements, whereas L. europaeus was the most distant. The results suggest that L. corsicanus and L. castroviejoi could have had a common ancestor occupying a large distribution area in SW Europe between Italy and Spain before the expansion of L. europaeus.

Key words. Taxonomy, systematics, hares, Lepus, Leporidae, Lagomorpha, mammals, Europe.

Introduction

Unlike in other Mediterranean countries, hare systematics in Italy have never been a controversial topic. Two species were commonly accepted as representing the hares of the Italian peninsula, Lepus europaeus meridiei Hilzheimer, 1906, described from Aveyron (S France) and Lepus corsicanus de Winton, 1898, described from Corsica. Shortly after the description of the two taxa, Miller (1912) identified both hares as subspecies of L. europaeus. According to this author, the northern Italian hares correspond to L. europaeus meridiei and the southern Italian hares to L. europaeus corsicanus. © Biodiversity Heritage Library, http://www.biodiversitylibrary.org/; www.zoologicalbulletin.de; www.biologiezentrum.at

60 F. Palacios

Miller's view accepting these hares only as different subspecies seems unusual for a period in which this author and many other taxonomists were very generous in recognizing the validity of hare species, and the decision seems a little incongruent in comparison with other taxonomic interpretations, as in the case of Lepus par- nassius Miller, 1903 from Greece, in which the holotype has the typical skin and skull of L. europaeus. However, in Italy Miller (1912) separated peninsular hares only at the subspecies level despite the fact that he clearly identified L. europaeus corsicanus on the basis of peculiarities of pelage color, and even after detecting the discriminant underfur color of L. corsicanus in specimens from Sicily and southern Italy. Petter (1961) in a wide revision of European and Asiatic hares classified both Italian subspecies under Lepus capensis, without making any particular reference to the hares from Italy. In a descriptive work on Italian hares, Toschi (1965) followed Miller (1912) and Ellerman & Morrison-Scott (1951) for taxonomy on peninsular and Sicilian populations. Following these papers nobody has paid further attention to the study of Italian hares and in the latest publications updating the main taxonomic problems still af- fecting Old World hares by Flux (1983), Angermann (1983), Corbet (1986), and Flux & Angermann (1990) there is no reference to them. However, in a recent review, Hoff- mann (1993) lists L. corsicanus as a valid species on the basis of a preliminary con- tribution on this same subject (Palacios et al. 1989).

Material and methods

I studied hare specimens from Italy held in the collections of the Istituto Nazionale di Biologia della Selvaggina "Alessandro Ghigi" (Bologne), Museo Civico di Storia Naturale (Milan), Museo Civico di Storia Naturale "Giacomo Doria"(Genova), Museo Zoológico de "La Specola" (Florence), Unidad de Zoología Aplicada (Madrid), Muséum National d'Histoire Naturelle (París), Natural History Museum (London), American Museum of Natural History (New York), and National Museum of Natural History (Washington). Data collected include label information, specimen identification, skull size and morphology, pattern of enamel in dentition, and pelage color and pattern. As the scope of the taxonomic problem to be clarified here is restricted to hares previously identified as brown hares (L. europaeus) from Italy and Corsica, I excluded Italian L. timidus and the hares from Sardinia that belong to L. mediterraneus. I also excluded brown hare specimens collected after the end of the 19th century to avoid the influence of hare restocking. As far as I know, this has been a common practice in Italy for hunting purposes since the early 1900s, according to label data for specimens from the MF collection, with an increase in this activity after 1930 (Toschi 1965). Sample limits set in this paper are of interest both to define the natural distribution area of the brown hare in Italy and to characterize the indigenous population. Identification of brown hare specimens has been done using previous experience on Spanish hares (Palacios 1989).

As for the Corsican hare (L. corsicanus) all available specimens from the Italian Peninsula, Sicily, Corsica, and Elba Island were included in the sample. It contains specimens from the 19th century and a few from this century. The identification of the specimens was based on the characteristics of the holotype which was captured in Corsica and acquired in the market of Bastia. As regards the origin of this island population, there is evidence to support the view that hares were introduced. Vigne (1988) indicates that Corsican hares were probably released on the island no later than in the 16th century. • Only relative age class IV adult specimens (Cabon-Raczynska 1964, Palacios 1989) were included in the comparison between samples. Some of the criteria used to identify the specimens with available skulls are: exooccipital-supraoccipital suture completely ossified and not detec- © Biodiversity Heritage Library, http://www.biodiversitylibrary.org/; www.zoologicalbulletin.de; www.biologiezentrum.at

Indigenous hares of Italy 61

table by color or thickness; mandibular capitulum well-ossified lacking porosity or fissures under the condyles; temporal tubercles and supraorbital processes well-developed; absence of porosity on the nasal lateral processes of the frontal bone and on the edges of the supraorbital processes. In specimens represented by skins only the relative age (adult, subadult, young) was estimated using other criteria based on the color and width of centrodorsal hair bands, body size (the dry measurements of hindfoot and ear were used when necessary), degree of ossifica- tion of long bones attached to skins, reproductive data, development of mammary nipples, etc. As some of the old specimens in Italian collections are mounted with skull inside, not all the skulls of the specimens listed below were available for study. The Milan and Genova collections do not hold either specimens of L. corsicanus or old specimens of L. europaeus, but they have been examined to complete the information on contemporary specimens of L. europaeus and other hare species from neighboring areas. The Madrid collection was used as a reference for data on the Spanish species. The L. corsicanus and L. europaeus specimens from Italy that were studied are listed in Ap- pendix 1 and 2, respectively. The available body measurements were taken from labels, and dry measurements were taken in some specimens on the stuffed skins with a digital calliper. Skull measurements were taken from all specimens using a standard digital calliper connected to a printer. Measurement abbreviations are listed in Appendix 3. To take some of the skull measurements calliper points were modified by sharpening them on both sides, and the longest ones were fluted transversely on the internal side near the far end; cranial foramina were measured using a set of wires of known diameter with an accuracy of 0.1 mm; external radius of curvature of the first upper incisor (II/) measurements were taken using a circle template with accuracy of 0.1 mm; alveolar position of the 11/ end was determined by transparency using a lateral spotted light in a dark room; drawings of teeth in cross-section were made with camera lucida using refracted light, first upper incisor (II/), second upper premolar (P2/), third upper premolar (P3/), and third lower premolar (P/3) were drawn at the occlusal surface, and 11/ also in a cross-section, internal to the alveolus, located % of the total curvature posteriorly. Data on pelage color are limited to a comparison of centrodorsal hair bands in all the specimens with the underfur and the dominant hair considered separately, and observations of interspecies color variation in a number of body regions in some specimens of each species; pelage color was studied mainly from winter specimens by comparison with Ridgway's (1912) color table. Univariate statistics for the body and skull variables were computed with Statgraphics and BMDP software. The Levene test was used to determine if the variances were equal or unequal and, in the case of unequal variances, the Welch correction was used to calculate the test of hypotheses regarding the differences between means. The Chisquare analyses made to test differences in frequencies of skull and dental characters between species were performed using BMDP software. The correspondence factor analyses made to represent graphically the differences between species in frequency data of skull and dental characters were carried out using BIOMECO software of the Centre Louis Emberger (CNRS, Montpellier). The stepwise discriminant analyses to calculate the taxonomic relationship among species using skull measurements were performed using BMDP software.

Results External measurements

The body size of indigenous Italian hares is not well-documented because most of the old museum specimens lack external measurements. On the other hand, the few specimens of relative age IV that have external measurements, particularly in the sample of L. corsicanus, show a great disparity in the figures for each measurement

(Table 1) that is hard to accept as intraspecies variation within Lepus. For instance, the tail length and hindfoot length have abnormally high variation coefficients, between 20 and 30 °fo. Figures for these coefficients in samples of other hare species © Biodiversity Heritage Library, http://www.biodiversitylibrary.org/; www.zoologicalbulletin.de; www.biologiezentrum.at

62 F. Palacios

1 Table : External measurements of some adult L. corsicanus and L. europaeus specimens from Italy taken from skin labels. (1) indicates that the figures correspond to "lungheza mas.". In the other specimens they appear under the terms "testa et corpo" or "head and body length". Description of measurements is in Appendix 3.

Specimens HB T HF E

L. corsicanus BM 8.9.30.1 610 70 90 105 AMNH 160956 550 109 141 94.1 AMNH 160959 590 120 126 96 L. europaeus BM 98.10.2.18 550 (1) 90 130 107 BM 98.10.2.19 500 (1) 90 140 107 USNM 153400 558 92 138 112

from Spain of relative age IV have never been higher than 12 % for tail length and 4 % for hindfoot length (Palacios 1989). The only explanation for this fact is that these measurements of Italian hares have been taken by each collector in a different way (tail length from base or from anus, with hair or without hair, or as the difference between total body length and head & body length; hindfoot with nail or without nail; ear from notch or from base). As for hindfoot length, a control of the figure validity for L. corsicanus specimen BM 8.9.30.1 was done using the dry hindfoot length taken on the stuffed skin. As the figure for the dry measurement without nail is 123 mm and the foot is straight and seems intact, the available specimen label figure of 90 mm is not possible. Other figures for dry hindfoot without nail in L. corsicanus are 118 and 120 mm in specimens BM 78.7.3.4 and BM 98.2.9.1, respectively. Considering that the hindfoot length without nail in hares undergoes an average decrease of 5.08 % from fresh to dry conditions (P <0.001, n = 33, sd = 1.54), it can be estimated that the fresh hindfoot length without nail of the three L. corsicanus specimens would range from

124 to 130 mm. On the basis of this estimate it is likely that the label figure of 141 mm available for the hindfoot length of specimen AMNH 160956 includes the nail. De Winton's (1898) dry hindfoot measurement of the L. corsicanus holotype (BM 78.7.3.4) is 119 mm. This figure is similar to the dry measurement reported above, but the dry ear measurement (107 mm) was probably taken from the base as the dry figures from the notch that I took on the skin are 84 and 90.5 mm (right and left ear). As regards L. europaeus from Italy, the size of the hindfoot seems somewhat larger than in L. corsicanus. The fresh hindfoot measurement of specimen USNM 153400 is 138 mm. This figure could be valid because the value for the dry measurement without nail is 134.8 mm. On the contrary, the figure of 130 mm corresponding to the specimen BM 98.10.2.18 seems somewhat low for relative age IV specimens of L. europaeus as we found equivalent figures in fresh specimens from Spain of relative age II (juveniles having the exoccipital-supraoccipital suture not yet ossified). As for ear length, the available label figures for L. corsicanus 105, 94.1 and 96 mm, corresponding to specimens BM 8.9.30.1, AMNH 160956 and AMNH 160959, respectively, are smaller than those for L. europaeus with figures of 107, 107 and 112 © Biodiversity Heritage Library, http://www.biodiversitylibrary.org/; www.zoologicalbulletin.de; www.biologiezentrum.at

Indigenous hares of Italy 63 mm corresponding to specimens BM 98.10.2.18, BM 98.10.2.19 and USNM 153400. This difference is also supported by the information derived from dry ear length figures taken from the notch in specimens of L. corsicanus. These measurements are 91, 84 and 90 mm in specimens BM 8.9.30.1, BM 78.7.3.4 and BM 98.2.9.1 and 84 mm in MNHN 1919-670. Considering that the ear length taken from the notch experiences a reduction of 10.2 % in hares from fresh to dry conditions (P < 0.001, n = 59, sd = 2.44), the fresh measurement for those specimens can be estimated at between 92 and 100 mm, a range that is still less than that of L. europaeus. The figures for head & body length of L. corsicanus (written on labels as "testa et corpo") seem excessively large and it is likely that they, in fact, correspond to the total length measured from the front of the snout to the end of the tail. This assump- tion is supported by comparison with L. europaeus, a larger species in which, according to my data on other European populations, the head & body length of fresh adult specimens rarely attains figures over 550 mm, with figures between 530 and 550 mm being the norm.

The head and body length figures for the Italian sample of L. europaeus are still difficult to interpret. It depends on the exact meaning of the term "lungheza mas", which is written on the labels of specimens BM 98.10.2.18 and BM 98.10.2.19 near the figures 550 and 500 mm. This term is probably the equivalent here to head & body length as the fresh measurements of total length in adult L europaeus are rarely below 600 mm. However, a priori the meaning of "lungheza mas" seems to be different from the meaning of "testa et corpo" which appears on the labels of the other specimen, USNM 153400, before the figure of 558 mm. In short, body size in L. corsicanus appears to be smaller than in L. europaeus, but interpretation of the scarce external measurements of the collection specimens entails some confusion, precluding their utilization for precise comparison between the two species. Body measurements of new L. corsicanus specimens would be of interest in the future for a full comparison with L. europaeus.

Skull measurements Skull size has been well-characterized on the basis of 29 measurements from two samples of ten specimens each of every species corresponding to relative age IV. These samples may be considered as few in number but are close to the maximum that can be gathered for study considering the specimens of indigenous hares of relative age IV from Italy (and collected during the 19th century) available worldwide. Tables 2 and 3 represent the values corresponding to the measurements of L. corsicanus and L. europaeus, respectively, and Table 4 depicts the results regarding univariate statistics.

L. corsicanus has a smaller skull than L. europaeus as is shown by most of the 29 measurements analyzed. In 9 of these measurements there is no range overlap between species. On the other hand, these measurements are very uniform and the values of the coefficients of variation are, in general, very small. In the case of the total length (TL) these coefficients are 2.9 °7o in L. corsicanus and 2.3 °7o in L. europaeus. The other variables have similar figures. For instance, as regards the palatal length (PL), they are 1.9 % and 4.0 °7o, respectively. The only one of these 9 variables with a large coefficient of variation is the height of processus muscularis mandibulae 8 4 3

64 F. Palacios

Table 2: Skull measurements of 10 adult specimens of L. corsicanus. 1 — MF 10870, 2 — MF 10871, 3 — MF 11526, 4 — MF 11588, 5 — BM 8.9.30.1, 6 — BM 78.7.3.4, 7 — BM 98.2.9.1, 8 — BM 19.7.7.2341, 9 — AMNH 160956, 10 — AMNH 160959. The complete data for these specimens is in Appendix 1 and the description of skull measurements is in Appendix 3. Negative values of DIRSIM when the incisor root does not reach the suture and the value 0 when the posterior end of the incisor root reaches the level of the suture.

Specimens

Variables 1 2 3 4 5 6 7 8 9 10

TL, 90.45 89 99 - 84.25 92 9 89.0 89-5 - 92.0 92.01

PL 35-8 36 5 37.12 34.87 36.3 36 1 35-6 37.0 36.23 37.18

INL 31.23 29 01 33.01 30.71 31 0 31 3 31.5 32.8 29.ll 31.52 ENL 40.31 38.47 40.22 39.51 40 2 39 6 40.6 39.1 38.01 40.86

UCTRL 17.06 16 73 17.4 17.12 17 1 17 0 18.0 17.O 16.87 17.48 FIL 22.36 23 15 23.67 22.68 23.0 23 0 22.75 24.0 22.61 23.09 SFW 13.24 13.87 13.91 15.35 13 2 14 6 12.5 14.4 15.16 12.67 WSP - " 29.93 30 8 31 8 31.0 30.8 - - ANW 13-75 12.58 12.74 12.66 14 05 13.1 14.7 12.0 13-39 14.

PNW 21.62 20 86 20.9 20.76 21 1 20 9 21.8 20.5 20.56 21.04

WFT 40.29 40 12 40.53 39.58 42 9 41 3 42.3 41.2 42.73 41.6 PZW 43.29 42.96 ~ 41.67 43.9 43.7 43.59 43.51

RW 25.17 25.49 26.26 25.15 26 9 26 0 26.3 26.7 27.33 26.77 FIW 9.66 11 02 10.93 11.21 11 35 10 75 11.25 11.6 10.72 9.52 NPB 6.87 6.76 7.17 6.22 6 75 6 6 5-9 7.0 7.1 7.43

10. 36 10 12 10 . 06 10. 62 11 45 10 3 10 . 10 . 9-57 10.18

FUISW 2.94 3 11 3.0 2.86 2 9 2 9 2.95 2.95 3.12 2.91

FUISL 1.95 1 9 1.82 1.78 1 85 1 65 1.8 1.7 1.93 1.84

RIC 9.6 10 7 10.6 10.0 10 2 10 8 11.5 10.3 10.3 10.7

FTL 7.85 7 59 7-97 8.56 8 2 7 8 8.8 9-7 9-79 8.0 TBL 10.84 10 13 11.23 11 15 10 2 11.0 10.66 11.19

TBW 7-79 7 39 6.58 7 5 7 0 7-6 8.03 8.11

DFZ 0 0 0.8 0.4 0 1 2 1.3 0.5 1.2 1.0 í

DPFP 0.3 0 3 0.5 0.3 0 4 0 2 0.4 0.6 0.6 0.4 MLCP 68.83 68 48 64.39 69-5 66 0 69.7 67-5 67-57

MH 40.45 39 68 37-12 40 4 37 9 40.2 39-5 40.74

LCTRL 18.27 17 92 18.16 18.39 17 8 17 9 19.0 18.15 17.48

HPMM 0.48 O.69 0.88 1 15 0 7 1.5 0.9 0.84 0.95

DIRSIM -2.62 -2 38 -1.43 0.0 -1 5 -0 9 -0.7 -1.0 -0.5 0.0

(HPMM) which is 32.5 % in L. corsicanus and 8.4 % in L. europaeus. This is because the figures for this variable have values which are of the same or less than the difference between min. and max. values, particularly in L. corsicanus (mean value 0.89; range 0.48 — 1.5). Fig. 2, g illustrates well the shape of processus muscularis mandibulae in the two species.

Fig. 1 illustrates fairly well the important distinction between the two taxa as regards skull measurements. It corresponds to a three-dimensional plot where variables without overlap (HPMM, MLCP and INL) have been represented. The differences in skull size are also backed by other variables which, although presenting some overlap, still differ significantly. Among them two variables are © Biodiversity Heritage Library, http://www.biodiversitylibrary.org/; www.zoologicalbulletin.de; www.biologiezentrum.at

Indigenous hares of Italy 65

Table 3: Skull measurements of 10 adult L. europaeus specimens. 1 — MF 10871 (dupl), 2 — MF 10872, 3 — MF 10874, 4 — MF 11523, 5 — MF 11525, 6 — MF 11527, 7 — BM 98.10.2.18, 8 — BM 98.10.2.19, 9 — BM 19.7.7.2486, 10 — USNM 153400. The complete data for these specimens is in Appendix 2 and the description of skull measurements is in Appendix 3.

Specimens

Variables 1 2 3 4 5 6 7 { ] 9 10

TL 97.47 102 2 96 42 96 49 98 43 97 29 95 31 97 08 95 29 100 8 P L 40.94 41 45 39 86 39 68 40 65 40 38 38 4 38 53 37 8 43 3

INL 38.73 35 16 42 74 37 97 39 55 33 5 37 09 38 27 34 11 40 29 ENL 44.92 42 16 45 36 44 6 47 81 41 44 44 75 45 03 41 65 47 21 UCTRL 17.81 18 46 18 83 18 1 17 38 17 23 16 27 17 96 17 73 18 33 FIL 26.36 26 68 24 4 25 48 26 75 25 29 26 2 25 79 24 08 28 30 SFW 13.45 16 0 14 51 16 07 13 75 14 87 14 76 13 9 14 32 13 2 WSP 32.92 35 55 39 01 34 1 34 91 34 46 34 95 34 69 36 34

ANW 15-39 15 72 16 15 14 5 14 93 13 98 17 02 14 86 16 56

PNW 21.92 22 35 21 2 23 99 22 53 22 21 22 16 23 8 22 3 25 1

WFT 41.19 43 85 44 37 43 25 43 19 44 02 42 06 4l 9 42 4 42 76 PZW 43-67 45 92 46 43 45 98 ^5 86 47 78 44 29 43 15 45 44 46 57 RW 27.39 28 43 29 36 26 13 27 38 26 82 26 67 27 66 27 88 28 34 FIW 11.46 11 37 11 91 12 14 11 97 11 14 11 98 12 33 11 58 13 29

NPB 6.61 7 0 6 63 6 14 6 26 7 21 5 82 5 01 5 8 6 05 PPW 12.06 10 78 12 01 11 62 11 32 11 24 11 57 11 52 12 03 13 09 FUISW 3-04 3 2 2 93 2 99 3 01 2 85 3 23 2 94 3 2

FUISL 2.16 2 21 2 1 1 91 1 98 1 91 2 13 2 06 2 22

RIC 9-4 10 9 11 0 10 1 9 8 9 7 9 8 10 7 9 8

FTL 8.24 7 94 8 59 7 92 9 2 7 65 7 67 8 76 8 05 8 17 TBL 712.26 12 31 11 17 11 24 11 67 11 84 12 16 12 03 10 92 11 0b TBW 8.85 8 86 8 57 8 09 8 37 8 75 8 45 8 93 8 92 8 41 DFZ 0.3 0 2 0 7 0 3 0 2 0 3 0 9 0 6 0 0 0 0

DPFP 0.8 0 8 0 9 0 9 0 8 0 8 0 6 0 8 0 7 0 8 MLCP 77 07 71 92 74 58 73 82 73 62 70 75 73 47 72 76 77 81 MH 43 33 41 16 41 64 41 02 41 3^ 42 63 44 17 ^3 05 45 11

LGTRL 18 28 19 48 19 17 18 79 17 5 18 56 17 65 18 99 HPMM 3 06 2 82 2 66 2 48 2 56 3 04 3 17 2 35 DIRSIM -4.83 -3 53 -2 32 -4 31 -4 22 -3 76 -2 5 -4 21 -2 49 -5 74

noticeable: the distance between the end of the root of the first upper incisor and sutura incisivom axillaris (DIRSIM) (Fig. 2, d) and the diameter of the posterior foramen palatinum (DPFP). In both cases the overlap is very small, particularly in the second case which is reduced to the minimum, the highest value of L. corsicanus being the same as the smallest of L. europaeus. In spite of the larger overall skull size for L. europaeus, there are a few variables which are larger in L. corsicanus than in L. europaeus. Among them, the narrowing of the palatine bridge (NPB), radius of the first upper incisGr curvature (RIC), facial tubercle length (FTL), and diameter of foramen zygomaticus (DFZ) are worthy of note. These are the variables showing negative percentages of difference between spe- . ;) \ ) )

66 F. Palacios

Table 4: Univariate results corresponding to skull measurements of adult L. corsicanus and L. europaeus. Values of mean, range, coefficient of variation within each species, and statistical t, significance level, and °7o of difference between species are given. Variables marked with (1) have unequal variances; ns = difference between means is not significant (P >0.05); * = somewhat significant (P <0.05); ** = significant (P <0.01); *** = very significant (P <0.001); (2) = P < 0.0001.

L. corsicanus (n=10) L. europaeus (n=10) sign.

Variable mean min-max cv t level % mean mm - max cv

/ft/i 9c no n\ *** c -6 • LKJi- <- 2 TL 90 01 2 9 . 57 (2) Q 3 97 67 \ y j ¿7 I 3

í o)i O7.07 -1 Q\ 1 -6.76 *** 10 40 .8-43 4 0 PL (1) 36 27 \ JT-ö/-j/ lo; 9 (2) 09 (37 . 3) *** lr>o i 9Q ni m S-42 74} 11 • ui- ui 4 1 -6.63 (2) 21 7 6 INL (1) 31 { ¿9 H ) J 37 »** (ill . 2 -6.42 12 44 .44-47.81) 4 8 ENL 39 68 ( 30 Ul -HV OD I 3 (2) 49

3 t 1 8^,) -2 . 7 81 6 27-1 8 4 0 UCTRL 17 17 ( id . /3-10 . u; 2 0 46 j 17 *** . (24 .08-28 4 FIL (1) 23 03 [ ¿l¿- . 3D ¿4 u 2 1 -6.90 (2) 12 25 93 . 35) 7

1 1 0 -1- 1 ik ns 14 2 14 48 ?-i f, 07 6 SFW 13 89 (12.5-15-35) 6 4 . 34 7 *** 1 14 2 M20<- - Q2-^Q7<- • 01)KJ± I 4 WSP (1) 30 86 (¿>7J J 1 -O) 2 0 • J¿ 35 27 V J 7 9

~> *** ANW 13 32 / 1 n- 1 ü 7 ^ 6 4 -4 QQ 15 q 15 45 (1^ 98-17 02) 6 4 *** / on c;-?i ftl .2-25 . . ¿-J. . -4 8 22 (21 0 PNW (1) 21 00 V ¿-W 7 0 ) 1 9 .46 3 75 1 5 42 (41.19-44.37) 2 WFT 41 25 \ Jy • 0° 4¿ • 7 ; 2 7 J • J 7 3 89 3

»#* c; iii-i / ii 1 £7_iio Q ^ 1 -3.86 2 1R-Ü7 78) 0 PZW (1) 43 23 (41.D/ -43 . y 7 _) 45 50 3 *# RW 26 20 (25.15-27.33) 2 8 -3-65 5 3 27 60 (26.13-29.36) 3 4

FIW 10 80 (9.52-11.6) 6 3 -3-82 10 2 11 91 (11. 14-13. 29) 5 0 NPB 6 78 (5.9-7.43) 6 6 1.77 ns -6 7 6 32 (5. 01-7. 21) 10 4 PPW 10 38 (9.57-11.45) 4 8 -5-28 (2) 12 9 11 72 (10. 78-13. 09) 5 2

FUISW (1) 2 96 (2.86-3.12) 2 7 -1.52 ns 2 7 3 04 (2.85-3.23) 4 2

FUISL 1 82 (1.65-1. 95) 4 9 -5-14 (2) 13 7 2 07 (1. 91-2. 22) 5 3 RIC 10 47 (9.6-11.5) 4 8 1-33 ns -3 2 10 13 (9.4-11.0) 5 7

FTL 8 42 (7.59-9-79) 9 2 0.71 ns -2 4 8 21 (7.63-9-2) 5 9

TBL 10 80 (IO.13-11.23) 3 9 -3.71 7 9 11 66 (10. 92-12. 3D 4 4 »»» TBW 7 50 (6.58-8. 11) 6 8 -5-85 (2) 14 9 8 62 (8.09-8.98) 3 3

DFZ 0 82 (0.0-1.3) 52 4 2.78 -57 3 0 35 (0.0-0. 9) 82 8

DPFP 0 40 (0.2-0.6) 32 5 -7-73 (2) 97 5 0 79 (O.6-O.9) 10 1

MLCP 67 74 (64.39-69-7) 2 6 -6.20 *# (2) 9 1 73 97 (70.75-77-81) 3 0

MH 39 49 (37-12-40.74) 3 2 -4.63 7 8 42 60 (4l.02-45.ll) 3 3 18 6 LCTRL 18 11 (17.48-19-0) 2 • 3 -2.07 ns 3 68 (17.5-19.48) 3

HPMM 0 89 (0.48-1.5) 32 -5 -14.49 (2) 217 9 2 83 (2.48-3-17) 8 .4 -1 -3 (-5.74/-2.32) .2 DIRSIM .10 (-2.62/0.0) 80 • 9 5-95 (2) 244 5 • 79 29

cies in Table 4. Other variables showing the same tendency in relative terms as the above ones, but with absolute figures slightly smaller in L. corsicanus than in L. europaeus, are the smallest frontal width (SFW), the first upper incisor section width (FUISW), and the lower cheek tooth row length (LCTRL). The variables showing the greatest percentage of difference between species are the DPFP, HPMM and DIRSIM.

Skull characters The skulls of L. europaeus and L. corsicanus are rather different in shape. L. europaeus generally has a massive skull with heavy structures while L. corsicanus has a light skull and the structures are more delicate than in the brown hare. The upper © Biodiversity Heritage Library, http://www.biodiversitylibrary.org/; www.zoologicalbulletin.de; www.biologiezentrum.at

Indigenous hares of Italy 67

contour of the skull from an lateral view in L. europaeus is straighter than in L. corsicanus which has a more rounded shape. This is mainly the effect of the nasals that are generally flat and long in L. europaeus and somewhat curved and shorter in L. corsicanus, as well as the brain case which in L. corsicanus is more rounded than in L. europaeus. However, most of the morphological difference between L. corsicanus and L. europaeus is the result of the influence of a number of skull structures related to 15 different characters, which are compared in the two species below. The character states presented by each specimen are shown in Tab. 5.

A) Position of the lateral foramen palatinum (Fig. 2, a, 1). — In L. corsicanus specimens these foramina are normally visible on the ventral plate of the palatine process of the maxilla, near the P2/ (state 1). Some specimens of L. corsicanus have a second pair of foramina occupying the position defined below as typical for L. europaeus, or in an intermediate site (state

2) between the position typical for L. europaeus and the position defined as state 1. Most of the specimens of L. europaeus have these small foramina on the anterior edge of the palatine process of the maxilla, positioned laterally (state 3). This site is internal to the incisive foramina and is not visible from the perpendicular to the skull in ventral position. L. europaeus specimens with a second pair of foramina near P2/ or in an intermediate position are not common. B) Position of the posterior foramen palatinum (Fig. 2, a, 2). — Most L. corsicanus specimens have these foramina totally enclosed in the lamina horizontalis of the palatine bone

(state 1) while in L. europaeus these foramina are positioned exactly on the suture separating the lamina horizontalis from the palatine process of the maxilla (state 3). A few specimens of L. corsicanus have these foramina in an intermediate position that is close to state 1, but with the lamina horizontalis opened in front to the suture (state 2). The lamina horizontalis is more extensive in L. corsicanus than in L. europaeus and in many specimens of L. corsicanus it is slightly pitted, this being another distinction from L. europaeus which never has a pitted lamina horizontalis.

C) Symphysis of the incisive bone (Fig. 2, h). — L. corsicanus specimens normally have a small but well-developed crest on the top of the anterior part of this symphysis (state 1), while © Biodiversity Heritage Library, http://www.biodiversitylibrary.org/; www.zoologicalbulletin.de; www.biologiezentrum.at

68 F. Palacios

1 cm

Fig. 2: Detail of several skull characters with different shapes in adult L. europaeus (left) and

L. corsicanus (right), a: 1 — position of the lateral palatal foramina, 2 — position of the posterior palatal foramina, 3 — upper edge of canalis zygomaticus, 4 — shape of the facial tubercle; b: shape of the fronto-nasal suture; c: 1 and 2 — shape of the oral and aboral part of the supraorbital process; d: position of the end of the first upper incisor root with respect to the incisivo-maxillaris suture; e: 1 — pitted degree in parietal surface, 2 — shape of the coronary suture, 3 — posterior contour of the parietal, 4 — parietal projection between temporal and occipital bones, 5 — parietal trench; f: 1 and 2 — posterior contour and lateral margins of the medial parietal part of squama ossis occipitalis; g: shape of processus muscularis mandibulae; h: symphysis of the incisive bone. Drawings of L. europaeus correspond to specimen USNM 153400 and drawings of L. corsicanus correspond to specimens BM 98.2.9.1 (a, d), BM 78.7.3.4 (b, e), BM 8.9.30.1 (c, f, g), and MNHN 1962-2546 (h). © Biodiversity Heritage Library, http://www.biodiversitylibrary.org/; www.zoologicalbulletin.de; www.biologiezentrum.at

Indigenous hares of Italy 69

Table 5 : Skull character states of adult L. corsicanus and L. europaeus specimens. Charac-

ters represented by uppercase letters and states (1, 2, 3) are described in the text.

L. corsicanus Characters

Specimens A B C D E F G H I J K L M N 0

MF 10870 1 2 1 1 1 1 1 1 1 1 1 1 1 2 MF 10871 1 3 1 1 1 1 1 2 3 1 1 1 2 1 MF 11526 1 1 1 \ 2 2 1 3 1

MF 11588 1 1 1 1 2 1 3 2 1 3 1 2 1 1

BM 8.9.30.1 1 2 1 \ 1 1 1 1 1 1 2 2 1 2

BM 78.7.3.4 1 1 1 1 1 1 1 1 1 1 2 1 1 1 1 BM 98.2.9.1 1 1 1 1 2 2 1 1 1 2 1 1 2 2 BM 19.7.7.2341 2 1 1 2 1 1 1 1 1 1

AMNH 160956 3 1 1 -L 2 1 1 2 1 2 1 1 2 1

AMNH 160959 1 1 1 1 1 1 1 1 1 2 1 1 1 1

L. europaeus Characters

Specimens A B c D E F G H I J K L M N 0

MF 10871 dupl 3 3 3 3 3 2 3 2 2 3 3 1 1 2 1 MF 10872 3 3 2 3 3 3 3 2 3 3 3 1 3 3 3 MF 10874 1 3 3 3 3 2 3 3 2 3 3 3 1 3 3 MF 11523 3 3 3 3 3 3 3 3 3 3 2 3 3 3 2 MF 11525 3 3 1 3 2 3 3 2 2 2 2 3 3 3 3 MF 11527 3 3 1 3 3 2 2 2 3 3 3 3 3 3 3 BM 98. 10. 2. 18 3 3 3 3 3 2 3 2 2 3 3 3 2 3 1 BM 98.10.2.19 3 3 3 3 3 3 3 3 2 3 3 1 3 2 2

BM 19.7.7.2486 3 3 3 3 2 3 3 2 3 3 2 1 3 2 1 USNM 153400 3 3 3 3 3 3 3 3 3 3 3 3 3 3 2

in L. europaeus this symphysis is normally smooth without a crest (state 3). The specimens having a slightly developed crest have been defined as state 2. D) Shape of the supraorbital processes. — In L. corsicanus the oral parts of these processes

(Fig. 2, c, 1) are generally short and parallel, being frequently fused with the frontal bone, and the aboral parts (Fig. 2, c, 2) are only somewhat divergent backwards (state 1); inL. europaeus the oral parts are larger than in L. corsicanus and are rarely fused with the frontal bone being frequently divergent forwards, and the aboral parts are greatly divergent backwards (state 2). E) Fronto-nasal suture (Fig. 2, b). — The shape of this suture in L. corsicanus is generally like a deep inverted open V with straight or slightly concave sides due to the fact that the lateral posterior ends of the nasals are usually sharp and because the frontal bone protrudes noti- ceably between the nasals (state 1); in L. europaeus the fronto-nasal suture has the shape of an open W in right position with more rounded posterior nasal ends and small frontal bone projection (state 3). The intermediate specimens have been characterized as state 2.

F) Parietal surface (Fig. 2, e, 1). — In L. corsicanus the parietal surface is normally very pitted (state 1), while in L. europaeus it is smooth or slightly pitted (state 3). The intermediate specimens have been characterized as state 2. G) Posterior contour of the medial parietal part of the squama ossis occipitalis (Fig. 2, f,

1). — In L. corsicanus there usually is a medial projection backwards (state 1) while in L. europaeus there is a medial notch, often with a slight pointed projection inside (state 3). The specimens with a straight contour have been typified as state 2. © Biodiversity Heritage Library, http://www.biodiversitylibrary.org/; www.zoologicalbulletin.de; www.biologiezentrum.at

70 F. Palacios

H) Lateral margins of the medial parietal part of the squama ossis occipitalis (Fig. 2, f, 2). — In L. corsicanus the lateral margins are rather parallel and anteriorly reach the parietals at the same level (state 1), while in L. europaeus the lateral margins diverge markedly forwards and in most cases the anterior ends remain at a lower level than the parietals (state 3). Speci- mens having parallel margins with anterior ends not reaching the parietals and terminating at a lower level, or having very divergent margins reaching the parietals anteriorly at the same level have been typified as state 2. I) Caudal groove of the coronary suture (Fig. 2, e, 2). — The groove that each parietal has anteriorly is generally very marked in L. corsicanus (state 1) and slight in L. europaeus (state 3). The intermediate cases have been typified as state 2. It is also noteworthy that in L. europaeus the anterior-internal part of the parietals protrudes more forwards into the frontal bone than in L. corsicanus. J) Posterior contour of the parietal (Fig. 2, e, 3). — In L. corsicanus the portion of the parietal contour that is in contact with the occipital is generally concave (state 1), while in L. europaeus this portion is usually straight (state 3). The intermediate cases have been typified as state 2. K) Parietal projection between temporal and occipital bones (Fig. 2, e, 4). — L. corsicanus usually has a narrow parietal projection separating the temporal and occipital bones (state 1) or a parietal entrant only (state 2); as regards L. europaeus only a few specimens display state 2, and the most common case is the occipital and the temporal bones in close connection without a projection or entrant between (state 3). L) Upper edge of canalis zygomaticus (Fig. 2, a, 3). — In L. corsicanus the upper edge of canalis zygomaticus is usually not prominent (state 1) or there is a slightly developed crest

(state 2), while in L. europaeus generally there is a markedly developed crest (state 3). M) Presence/absence of lateral foramina on the zygomatic process of the maxilla. — In L. corsicanus there usually is a small foramen above the crista facialis, anterior to the large zygomatic foramen (state 1), but in most specimens of L. europaeus there is no foramen (state 3); the specimens with diminutive cavities but no real foramen have been typified as state 2. N) Postero-external parietal furrow (Fig. 2, e, 5). — In L. europaeus there usually is a well- marked lateral longitudinal furrow on the parietal surface close to the junction with the occipital and temporal bones (state 3). This furrow is usually absent (state 1) or slightly marked (state 2) in L. corsicanus. O) Shape of the incisive foramina. — In L. corsicanus the incisive foramina usually have divergent margins along the first % of their length, and parallel lateral margins along the posterior Vz (state 1); this shape is basically similar in L. europaeus except that in this species 2 there is usually a subterminal enlargement approximately h of the incisive foramina length posteriorly, the foramina being wider at this point than at the posterior end (state 3). Speci- mens with only a slight enlargement have been typified as state 2.

Among the other noticeable features distinguishing the two species, the shape of the facial tubercles (Fig. 2, a, 4) is worthy of note. They are more divergent posteriorly protruding more outwards in L. corsicanus than in L. europaeus in which they are fairly parallel. The fact that the facial tubercles are more external in L. corsicanus than in L. europaeus is supported by values of the TL/WFT index for which the means, 2.17 in L. corsicanus and 2.26 in L. europaeus, show a fairly significant difference (P <0.01). The length of the facial tubercle also presents values that are relatively larger in L. corsicanus than in L. europaeus (means of TL/FTL index 10.87 and 11.91, respectively, P<0.05). The statistical comparison of the frequencies of the character states in the two spe- 2 cies is presented in Table 6. Figures of X (Chisquare) and the significance level are given. High values of X 2 and small probabilities in 8 characters suggest that the frequencies corresponding to the two species make a significant difference between © Biodiversity Heritage Library, http://www.biodiversitylibrary.org/; www.zoologicalbulletin.de; www.biologiezentrum.at

Indigenous hares of Italy 71

Table 6: Statistical comparison of frequencies corresponding to the skull character states of L. corsicanus and L. europaeus. Frequencies for each character can be easily obtained in Table 5.

X 2

A 12.844 ** B 16.364 *** C 13.333 *** *** D 19.000 (2) E 15.200 F 13.652 *** G 14.936 *** H 14.760 *** ** I 11.111 J 14.355 * * * K 7.969 * L 8.923 * M 8.100 * N 11.925 ** O 5.600 ns

them. The most remarkable differences lie in the position of the posterior foramen palatinum and the shape of the supraorbital processes. The results of a factor correspondence analysis based on the frequencies corresponding to the 15 skull characters in each species are plotted in Fig. 3. This plot corresponds to the projection on axes 1 and 2 that displays the maximum difference between species. The projections belonging to the specimens are enclosed in the contour line and indicate that the two taxa are separated without overlap. Axis 1 explains

36.08 °7o of the variance and contributes most to species differentiation, character states Dl, D3, HI, Jl, Fl, Gl, and G3 being, in that order, the most important. Axis 2 only explains 9.64 % of the variance, E2, 02, B2, L2, J2, Ol, and H3 being the most important character states. States 1 are typical of L. corsicanus, states 3 are typical of L. europaeus, and states 2 have small frequencies shared by the two species or appearing exclusively in only one species. According to Fig. 3 the most typical specimens in the case of L. corsicanus are ch, ce, ca, cj, and cf, and eb, ej, ee, ef, and ed in the case of L. europaeus. All these specimens are the most distant on the plot and are projected near the states that are more important in the separation, particularly those related to axis 1.

Dental characters The morphological pattern of enamel in the dentition of L. europaeus and L. corsicanus was studied on the basis of the characters of II/, P2/, P3/ and P/3 cross-sections. The general characteristics of these teeth are rather similar in the two species as is the rule among species of the genus. 11/ is rectangular with a shallow anterior sub- central simple groove separating two lobes which have a circular contour, the lingual one being only somewhat narrower than the labial one. P2/ has a triangular-like shape with three main anterior grooves, being from right to left in a right tooth mesoflexus, paraflexus (the deepest one), and hypoflexus. These main entrants are, respec- © Biodiversity Heritage Library, http://www.biodiversitylibrary.org/; www.zoologicalbulletin.de; www.biologiezentrum.at

72 R Palacios

Fig. 3: Projection of the frequencies corresponding to the skull character states of L. corsicanus and L. europaeus specimens in relation to axes 1 (vertical) and 2 (horizontal) of the correspondence factor analysis. Couples of lower case letters represent the specimens of each species (e = L. europaeus; c = L. corsicanus; a, b, c . . . = same specimens and order as in Table 5) and upper case letters followed by numbers 1, 2, 3 represent the skull character states.

tively, between postcone and lagicone, lagicone and mesial hypercone, and mesial hypercone and distal hypercone. P3/ is also rectangular with a noteworthy transver- sal groove, the hypoflexus, that is open on the lingual side and crosses the tooth from one side to the other nearly reaching the labial wall. The hypoflexus has a fluted enamel structure especially on the anterior edge. Finally, P/3 has a subcircular shape with a deep anteroflexid separating anteriorly the two anteroconids, a slight but wide protoflexid on the labial side separating the labial anteroconid of the protoconid, and a long tranversal hypoflexid open to the labial side that nearly reaches the lingual © Biodiversity Heritage Library, http://www.biodiversitylibrary.org/; www.zoologicalbulletin.de; www.biologiezentrum.at

Indigenous hares of Italy 73

Fig. 4: Posterior cross section of several first upper incisors of adult L. europaeus (a) and L. corsicanus (b). a: 1 — MZ 10871 (dupl), 2 — MZ 10873; 3 — MZ 10874, 4 — MZ 11523, 5 — MZ 11524, 6 — MZ 11525, 7 — MZ 11527, 8 — MZ 11597, 9 — MZ 11598; b: 1 — BM 78.7.3.4, 2 — MZ 10870, 3 — MZ 10871, 4 — MZ 11526, 5 — MZ 11588, 6 — INBS 3830, 7 — AMNH 160956, 8 — AMNH 160959. The arrow points to the posterior contour of the section which has a different shape in each species.

wall and separates the protoconid from the hypoconid. The lingual and posterior margins of this tooth are normally rounded, lacking any entrant except for the small paraflexid that is located in antero-lingual position. Despite the general resemblance of L. europaeus to L. corsicanus, there are some dental characters that allow a distinction to be made. Below I describe seven characters. Specimen data are summarized in Table 7.

A) Posterior contour of the 11/ cross-section (Fig. 4). — The shape of this tooth determines one of the biggest morphological differences between L. europaeus and L. corsicanus and also enables us to distinguish between L. corsicanus and many other hare species. In L. corsicanus © Biodiversity Heritage Library, http://www.biodiversitylibrary.org/; www.zoologicalbulletin.de; www.biologiezentrum.at

74 R Palacios

Table 7 : Dental character states of adult L. corsicanus and L. europaeus specimens. Charac- ters represented by uppercase letters and states 1, 2, 3 are described in the text; columns r and

1 correspond to the right and left teeth. Asterisks indicate specimens belonging to relative age class III which have been added to increase the samples in both species.

L. corsicanus Characters

A B C D E F G Specimens r 1 r 1 r 1 r 1 r 1 r 1 r 1

MF 10870 2 2 1 1 1 1 1 1 1 1 1 1 1

virMF j.1 \j0R71o / x ]_ 1 2_ 1 1 1 ]_ 2_ 2_ MF 11526 1 1 1 1 1 1 1 1 1 1 1 2 1 MF 11588 1 1 1 1 1 1 1 1 2 1 1 1 1 BM 8.9.30.1 3 2 1 1 1 1 3 BM 78.7.3.4 1 1 1 1 1 _ 1 1 2 1 1 1 1 1 BM 98.2.9.1 - - 1 1 - 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 3 3

&MMT4 1 AnQ^^ 2 2 1 *

BMMT4 1 fiflQ^Q 1 1 2 2 2 TMDC * ^P^n 1 1 1 1 ]_

Characters

A B C D E F G Specimens r r r 1 r 1 r r r 1

MF 10871 dupl 3 3 3 2 1

MF 10872 1 3 3 _ 3 3 1 1 2 2 1 1 MF 10874 3 3 3 3 3 3 3 3 3 1 1 1 1 MF 11523 3 3 3 3 1 1 2 2 2 2 1 3 1 MF 11525 3 3 3 3 1 3 3 3 3 3 3 3 3 MF 11527 3 3 3 3 3 1 1 2 2 3 3 2 1 BM 98. 10.2. 18 3 3 3 3 3 2 3 1 2 3 3 BM 98.10.2.19 3 3 2 2 2 1 2 1 2 1 1 BM 19.7.7.2486 3 3 1 1 1 3 2 1 1 1 1 USNM 153400 3 3 3 1 1 3 3 1 2 3 3 MF 10873 * 3 3 2 3 2 1 2 3 2 3 3 MF 11524 * 3 3 3 3 1 3 2 1 3

MF 11597 * 3 3 2 2 1 3 3 •1 3 3 3 3 3 MF 11598 * 3 3 3 3 3 3 3 1 1 3 2 2 2

the posterior contour of the cross-section of this tooth is generally flat or slightly convex (state

1) while in L. europaeus it is always concave (state 3). The slightly concave morphotype (state 2) is very rare in L. corsicanus. B) Size of postcone and lagicone of P2/ (Fig. 5, a). — This character represents another important difference between the two species. In most specimens of L. corsicanus the postcone protrudes farther than the lagicone (state 1) while in specimens of L. europaeus the postcone usually protrudes less than the lagicone (state 3). State 2 representing specimens of both species in which both lagicone and postcone protrude to the same degree is very rare. C) Crenulation on the posterior edge of the P3/ hypoflexus (Fig. 5, b). — In L. europaeus there frequently is a deep and oblique entrant on the posterior edge of the P3/ hypoflexus, © Biodiversity Heritage Library, http://www.biodiversitylibrary.org/; www.zoologicalbulletin.de; www.biologiezentrum.at

Indigenous hares of Italy 75

1 mm

Fig. 5: Details of the occlusal surface of the second upper premolar (P2/), third upper premolar (P3/), and third lower premolar (P/3) of adult specimens of L. europaeus and L. corsicanus showing some dental characters of taxonomic relevance, a: P2/ (right side) P = postcone, L = lagicone, M = mexoflexus; b = P3/ (right side) the arrow shows the crenulation of the posterior edge of the P3/ hypoflexus; c: P/3 (left side) 1 = centroflexid on the anterior edge of the P/3 hypoflexid, 2 — slight crenulation on the posterior edge of the P/3 hypoflexid, 3 — crenulation on the internal side of the P/3 hypoconid, 4 — paraflexid on the antero-lingual border of P/3. Drawings of L. europaeus (left) correspond to specimens MZ 11525, USNM 153400, MZ 11525, and drawings of L. corsicanus (right) correspond to specimens MZ 11588, MZ 11526, and BM 78.7.3.4, from above to below, respectively.

oriented through the opening of the groove (state 3), while this deep crenulation was not found

in L. corsicanus (state 1). Two specimens of L. europaeus: show a slight crenulation (state 2).

D) Centroflexid on the anterior edge of the P/3 hypoflexid (Fig. 5, c, 1). — L. europaeus sometimes has a large centroflexid (state 3) and normally a smaller but differentiated centroflexid (state 2) occupying a medial position on the anterior edge of the P/3 hypoflexid. This groove appears to be a different structure from the typical inflexion shown by hares resulting from the primitive connection between trigonid and talonid. In most cases, L. corsicanus lacks

the centroflexid (state 1), while state 3 was not found in this species.

E) Slight crenulation on the posterior edge of the P/3 hypoflexid (Fig. 5, c, 2). — In L. europaeus there frequently is a small but well-defined crenulation occupying a central position on the posterior edge of the P/3 hypoflexid (state 3) while in L. corsicanus this crenulation

is normally absent (state 1), being slightly apparent (state 2) in very few cases. © Biodiversity Heritage Library, http://www.biodiversitylibrary.org/; www.zoologicalbulletin.de; www.biologiezentrum.at

76 F. Palacios

F) Crenulation on the internal side of the P/3 hypoconid (Fig. 5, c, 3). — L. europaeus may exhibit a deep and oblique entrance on the internal side of the P/3 hypoconid (state 3), oriented to the hypoflexid opening. This entrance is normally absent in L. corsicanus (state

1) and, if present, is never deep (state 2). G) Paraflexid or parafosetid on the antero-lingual border of P/3 (Fig. 5, c, 4). — The frequency in L. europaeus of a deep paraflexid or parafosetid in P/3 (state 3) was double that in L. corsicanus which lacks the paraflexid in most cases (state 1). Cases of slight paraflexid (state 2) were rare in both species. Another morphological difference found between the two species as regards dentition is also related to the shape of the 11/ cross section. In L. europaeus the shape is more square while in L. corsicanus it is more rectangular (Fig. 4). The difference is due to the fact that L. corsicanus has relatively larger measurements of first upper incisor section width (FUISW) than L. europaeus, while the figures for section length

(FUISL) are correlated with skull size. This point is supported by the data in Table

4 showing that there is a strong overlap in FUISW values while the means for FUISL differ very significantly. The statistical results of the comparison of the frequencies of dental characters for each species are presented in Table 8 where figures of X2 and significance level are given. Two of the seven characters studied have high X 2 values and low probabilities suggesting that the frequencies corresponding to the two species make a significant difference between them. These characters are the posterior contour of the 11/ cross section and the size relationship between postcone and lagicone of P2/. The results of a factor correspondence analysis based on the frequencies table corresponding to the 7 dental characters in each species are plotted in Fig. 6. This plot corresponds to the projections on axes 1 and 2 that display the maximum difference between species. The projections belonging to the specimens are enclosed in the contour line and it can be seen that the two taxa are separated without overlap. Axis

1 explains 30.78 % of the variance and contributes most to species differentation, character states Bl, Al, A3, D3, Fl, and B3 being the most important in that order. Axis 2 explains 16.88 % of the variance, B2, C2, F3, F2, D2, E3, and E2 being the most important character states. States 1 are typical of L. corsicanus, states 3 are typical of L. europaeus, and states 2 are those with small frequencies shared by the two species or appearing exclusively in only one species. According to Fig. 6 the most typical specimens of each species are ci, ca, cb, and eg in the case of L. corsicanus

Table 8: Statistical comparison of frequencies corresponding to the dental character states of L. corsicanus and L. europaeus. Frequencies used for the analysis correspond to columns

Ar, Bl, Cr + 1, Dr, Er, Fr, and Gr, which have the minimum number of missing data.

Character X 2 Significance level

A 17.000 *** B 21.315 (2) C 9.244 ** D 10.478 ** E 5.682 ns F 5.400 ns G 4.308 ns © Biodiversity Heritage Library, http://www.biodiversitylibrary.org/; www.zoologicalbulletin.de; www.biologiezentrum.at

Indigenous hares of Italy 77

Fig. 6: Projection of the frequencies corresponding to the dental character states of L. corsicanus and L. europaeus specimens in relation to axes 1 (vertical) and 2 (horizontal) of the correspondence factor analysis. Couples of lower case letters represent the specimens of each species (e = L. europaeus; c = L. corsicanus; a, b, c . . . = same specimens and order as in Table 7) and upper case letters followed by numbers 1, 2, 3 represent the projections of the dental character states.

and ek, ea, em, ee, and eg in the case of L. europaeus. All these specimens are the most distant in the plot and are projected near the more important states in the separation, particularly those related to axis 1.

Pelage characters L. europaeus and L. corsicanus differ in several pelage characters of taxonomic im- portance although the overall appearance of these two species seems fairly similar.

Information regarding this chapter is complete only for the color of the centrodorsal hair for which the data were taken for the whole set of specimens of the two species.

For the other characters the analysis is based on data of some specimens chosen as representative of those belonging to each species. © Biodiversity Heritage Library, http://www.biodiversitylibrary.org/; www.zoologicalbulletin.de; www.biologiezentrum.at

78 F. Palacios

L.co

Fig. 7: Position of different winter pelage characters of taxonomic relevance to distinguish between adults of L. corsicanus and L. europaeus. 1 — centrodorsal hair, 2 — nape, 3 — rump, 4 — facial band, 5 — subocular patch + whiskers patch, 6 — collar, 7 — inner surface of forelimbs, 8 — ventral white, 9 — transitional fringe between back and belly. See comments in text.

A) Centrodorsal hair (Fig. 7, 1). — As regards the color of the centrodorsal hair, there are differences in the dominant hair bands as well as in the underfur bands, but the dominant hair has some variation in L. corsicanus and only the underfur color provided definitive con- clusions on the species distinctiveness.

1) Dominant hair. — As indicated above, analysis of dominant hair needs further study but there are some differences which deserve to be typified now. The adults of the two species are very similar with respect to the length and the color of the bands, except at the base of the hair. In the upper part of the hair there is a black tip which is common to the two species; below there is a subterminal ring that is pinkish buff in L. corsicanus and cream buff in L. europaeus, and then there is a wide fringe of black color occupying the middle part of the hair, which is also present in the two species. Under this black central zone, in L. corsicanus there is a narrow pinkish buff band and below is the base of the hair which is white, sometimes with a slight brown tone. The last two bands are present on L. corsicanus specimens MF 11526, MF 11590, BM 78.7.3.4, BM 98.2.9.1, AMNH 160959, MNHN 1919-670. The specimens belonging to L. europaeus always have a long grey base under the central black band. However, the dominant hair has some variability in L. corsicanus, and some specimens have a pale grey base (MF 10870, MF 11588, MF 10871, MF 11592 and AMNH 160956). It is likely © Biodiversity Heritage Library, http://www.biodiversitylibrary.org/; www.zoologicalbulletin.de; www.biologiezentrum.at

Indigenous hares of Italy 79

L. corsicanus L. euwpaeus

Black Black

Pinkish buff Pinkish buff

Grey White

Fig. 8: Drawing that illustrates the color characteristics of the underfur of L. corsicanus and L. euwpaeus adults in winter pelage. The color of the base is a discriminant character between the two species.

that there is a moulting, seasonal or age influence on this character. It may also be a poly- morphic character. Thus, the current state of knowledge does not allow species separation in all specimens. 2) Underfur (Fig. 8). — The adult specimens of L. corsicanus have three well-differentiated bands, the grey basal one, the intermediate pinkish buff band with some cinnamon-buff effect, and the black upper band. All L. corsicanus specimens conform to this description which can be checked in specimens MZ 10870, MF 10871, MF 11526, MF 11588, MF 11590, MF 11592, BM 78.7.3.4, BM 98.2.9.1, MNHN 1919-670, AMNH 160956, and AMNH 160959. As regards the underfur color of L. euwpaeus, most of the adult specimens have two well- defined bands, the white basal one and the upper band which is black. In a few specimens there is a third band positioned between the other two but hardly detectable. It is of the same color as the medial band of L. corsicanus but is less marked than in the latter species. The underfur color description of L. euwpaeus can be checked in specimens MZ 10872, MF 10874, MF 11585, MF 11586, MF 11591, MF 11594, MF 11595, BM 98.10.2.18, BM 98.10.2.19, and USNM 153400. The L. euwpaeus specimens with a somewhat perceptible medial pinkish buff band are MF 11585, MF 11586. It is likely that the old skins of L. euwpaeus may have partially or totally lost the color of the intermediate band of the underfur which is normally rich in color in fresh specimens. However, there could be a big difference between L. corsicanus and L. euwpaeus as regards this intermediate band. Comparing old skins of the two species in the MZ collection stored under the same conditions for over one hundred years, all the specimens belonging to L. corsicanus still display a wide and richly colored band while in L. euwpaeus this band is scarcely perceptible or absent. Looking at the base of the centrodorsal hair of adult specimens, in L. euwpaeus the contrast between the white base of the underfur and the grey base of the dominant hair is noteworthy, and in L. corsicanus the contrast between the grey base of the underfur and the white base of the dominant hair, except in specimens of L. corsicanus with a grey base in the dominant hair. However, the color of the base of the underfur is a stable and discriminant character and, consequently, very useful for differentiating between adult specimens of L. euwpaeus and L. corsicanus. Juvenile L. euwpaeus also have a grey base in the underfur, as is the case, for instance, in specimens MF 11584, MF 11587. The grey color must turn to white at an early stage in L. euwpaeus. The specimen USNM 153399, belonging to relative age III, already has a white base.

B) Nape (Fig. 7, 2). — There is a very noticeable difference between the two species as regards nape color bands. In L. corsicanus the hair on the nape has a wide blackish base, for instance in the holotype specimen BM 78.7.3.4, while in L. euwpaeus the basal fringe is short and grey, for example in specimen USNM 153400. © Biodiversity Heritage Library, http://www.biodiversitylibrary.org/; www.zoologicalbulletin.de; www.biologiezentrum.at

80 F. Palacios

C) Rump (Fig. 7, 3). — In L. europaeus there is a conspicuous patch of grey over the rump in winter pelage. This patch is due to the mixed effect of the subterminal and terminal rings of the dominant hair which are white and black, respectively. L. corsicanus does not have this conspicuous patch because the subterminal ring of the dominant hair over the rump does not turn to white in winter as in L. europaeus. On the contrary, it remains nearly the same color as the rest of the dominant hair of the dorsal pelage. In L. corsicanus the rump color is normally somewhat lighter than the back as in other hare species but this is because the terminal and medial bands of the dominant hair and the upper band of the underfur have less black or are not so dark in the rump region as in the center of the back where they are intensely black. Specimens AMNH 190956, AMNH 160959 of L. corsicanus in winter pelage display an almost uniform back with a rump hardly distinguishable by the change of color from the rest of the upper parts while specimen MF 11586 belonging to L. europaeus displays a noticable grey patch on the rump. D) Facial bands (Fig. 7, 4). — L. corsicanus usually shows a greyish white facial band on both sides of the head in winter pelage, extending from the preocular patch to the lower base of the ears, over the lower cheeks. Specimens MNHN 1919-670, MF 11592, BM 78.7.3.4, BM 8.9.30.1, BM 98.2.9.1, BM 98.2.9.2 are normal representatives of this character. These facial bands are not present in most specimens of L. europaeus. In the Italian sample only specimen MF 11586 with strongly marked winter pelage as revealed by the conspicuous grey patch on the rump has a kind of facial band but not as extensive and with the same pattern as in L. corsicanus.

E) Subocular patches (Fig. 7, 5). — The Corsican hare displays a conspicuous ochraceous- tawny patch extending underneath and behind the eye which is more obvious in winter pelage

because of the marked contrast with the white facial patch; it is not present in L. europaeus. This character was already noticed by de Winton (1898) as one of the peculiarities of L. corsicanus. Specimens BM 8.9.30.1 and MNHN 1919-670 of L. corsicanus are very representative as regards this character. A smaller patch of the same color is also present in the whisker area of L. corsicanus.

F) Collar (Fig. 7, 6). — The collar is very uniform in color in the two species, having a marked vinaceous tonality in L. europaeus which varies between light vinaceous cinnamon and cinnamon-buff, and a marked buff tonality in L. corsicanus, which varies from pinkish buff to cinnamon-buff. Representative specimens are MF 10874, MF 11593, MF 11594 of L. europaeus and BM 98.2.9.1, BM 13.1.900, BM 78.7.3.4 of L. corsicanus. G) Inner surface of forelimbs (Fig. 7, 7). — In L. corsicanus there is a well-marked light buff area in the inner side of the forelimb contrasting with the clay color of the outer surface, while in L. europaeus the inner buff area is less marked and the color of the forelimb is almost uniform clay-sorrel.

H) White hair extension in underparts (Fig. 7, 8). — The area of white pelage on the under- pays is more extensive in L. corsicanus than in L. europaeus, basically because L. corsicanus lacks the transitional fringe between dorsal and ventral pelage that is typical in L. europaeus. Specimens BM 78.7.3.4, BM 8.9.30.1, BM 98.2.9.1, BM 98.2.9.2, MF 10870, MF 11592 are representative examples of L. corsicanus.

I) Transitional pelage between back and belly (Fig. 7,9). — In L. europaeus there is always a transitional fringe on the sides separating the dorsal and ventral pelage. It is very uniform in color, varying between light vinaceous cinnamon and vinaceous cinnamon. In L. europaeus specimen USNM 153400 this character is very well represented. As stated previously, L. corsicanus lacks this band and presents a marked contrast between the dorsal and ventral pelage. Specimens quoted in the previous section are also typical examples of this character in L. corsicanus.

Geographic distribution Fig. 9 shows the geographic localities appertaining to all the studied specimens for L. europaeus and L. corsicanus. The localities of L. europaeus show that up to the end of the 19th century this species occupied the northern part of the Italian Penin- © Biodiversity Heritage Library, http://www.biodiversitylibrary.org/; www.zoologicalbulletin.de; www.biologiezentrum.at

Indigenous hares of Italy 81

Fig. 9: Evidence on the pattern of distribution of indigenous hares in Italy based on localities of museum specimens. Localities of L. europaeus (black squares) correspond to specimens collected in the 19th century, and localities of L. corsicanus (black triangles) correspond to all the available specimens in collections, some of them belonging to this century. 1 — surroundings of Rome, 2 — Lago Fucino, 3 — Monti di Gargano, Foggia, 4 — Mongiana, 5 — Saponara, 6 — Vicari, 7 — Marsala, 8 — Isola d'Elba, 9 — Bastia, 10 — Aleria, 11 — San Casciano di Bagni, 12 — Siena, 13 — Arezzo (Alpe de la Luna), 14 — Gabbiano, Camugliano, Bientina, 15 — Rassina, 16 — Maiano, Firenze (Florence), 17 — Genova, 18 — Domodossola, 19 — Trieste, 20 — Torino (Turin).

sula and the localities of L. corsicanus indicate that this species occupied the southern part of the Italian Peninsula, Sicily, Corsica and Elba Island. © Biodiversity Heritage Library, http://www.biodiversitylibrary.org/; www.zoologicalbulletin.de; www.biologiezentrum.at

82 F. Palacios

It is likely that this parapatric pattern corresponds to the natural distribution of the species, with probably a contact zone existing between Siena and Rome. These localities are the most southerly and the most northerly points, respectively, of L. europaeus and L. corsicanus. The Corsican hare appears to be originally from the southern part of the Italian Peninsula where this species probably was geographically confined after the expansion of L. europaeus throughout Europe. The distribution of L. corsicanus in Sicily could be the result of introduced hares as is the case in Corsica where, according to Vigne (1988), hares were released no later than in the 16th century. This could also be the case of Elba Island from where the only known specimens are to be found in the Museum La Specola in Florence (MF). Current L. europaeus distribution in Italy does not conform to the natural situation because of artificial introductions carried out throughout Italy from the beginn- ing of the 20th century. The present distribution of L. corsicanus in Italy is not well known and this species could be in frank regression as a result of hunting and habitat reduction by L. europaeus introduced from other European countries. The capture of two L. corsicanus specimens in Catanzaro in 1974 and 1975 (INBS collection) sug- gests that this species could still exist in some mountainous parts of southern Italy. In Corsica, Vigne (1988) reports the decline of L. corsicanus, also due to hunting and the introduction of L. europaeus.

Phenetic relationships amongst SW European hares To assess the phenetic relationships existing among L. corsicanus, L. europaeus, L. castroviejoi and L. granatensis a stepwise discriminant analysis using the 29 skull variables was carried out. The L. europaeus and L. corsicanus specimens are those from Italy. The L. granatensis specimens were caught in Palencia and Cádiz provinces while the L. castroviejoi specimens were caught in Palencia, León and Asturias provinces, all in Spain. The analysis was checked to determine if the groups were different. The MANOVA test shows an F value of 33.42 (P <0.0001) which means that the differences among groups are very significant. Missing values were estimated and included in a analysis after making a stepwise estimation within species which provides a value for each variable case using only the correlated variables. The analysis was able to produce a discriminant function with 5 of the 29 variables

(MMPH, DFZ, TBW, WSP, and TL). The classification matrix indicates that 100 °7o of the specimens were classified correctly into the groups according to the classification function. However, the jacknife classification matrix indicates that 97.6 °/o of

Table 9: F values computed from the Mahalanobis D2 statistics that test the equality of group means for each pair of groups in the stepwise discriminant analysis based on skull size variables. Group means are plotted in Fig. 10.

L. corsicanus L. europaeus L. castroviejoi

L. europaeus 56.02 L. castroviejoi 14.97 45.31 L. granatensis 34.98 29.93 42.74 A

Indigenous hares of Italy 83

6. + c A N - 0 y B N - BB \ 3 ' c - / \ - ^ A / 2BB \ / C C L - / B B B CC 3 C / / ^C-C V - B B A 0. + ^AA— R - * A / 1 - \ 1 A A - . . V L E -3. + / 4 DD\ A" D .D

-6.

-9-

-6 -4

CANONICAL VARIABLE

Fig. 10: Scatter plot for the first two canonical variables of the stepwise discriminant analysis carried out using skull variables of SW European hares. Group A represents specimens of

L. corsicanus, B of L. europaeus, C of L. castroviejoi, and D of L. granatensis. 1, 2, 3, and 4 are the group means on which the test of equality showed in Table 9 is based.

the specimens were classified correctly into the groups, 100 % of L. europaeus, L. castroviejoi and L. granatensis, and 90 % of L. corsicanus, because one specimen of this species was mistaken for L. castroviejoi. Dispersion scattergrams for the first two canonical variables of the discriminant- function analysis (Fig. 10) showed no overlapping in the phenotypes of the four species. The total variance explained by the first two canonical variables was 90.8 °7o.

Canonical variable 1 explains 61.6 °7o of the variance and separates L. corsicanus and L. castroviejoi from L. europaeus and L. granatensis. Canonical variable 2 explains 29.2 % of the variance and separates L. granatensis and L. corsicanus from L. castroviejoi and L. europaeus. Table 9 shows the test of equality of group means for each pair of groups. L. europaeus—L. corsicanus are the most differentiated species with a value of 56.02. © Biodiversity Heritage Library, http://www.biodiversitylibrary.org/; www.zoologicalbulletin.de; www.biologiezentrum.at

84 F. Palacios

Fig. 11: Camera lucida drawings of occlusal sections of P/3 of several adult specimens of L. castroviejoi. Al and A2 — UZA 1983.10.24.02 (right and left), Bl and B2 — UZA 1983.10.28.01, CI and C2 — UZA 1983.10.28.02, Dl and D2 — UZA 1983.10.28.03, El and E2 — UZA 1983.11.16.01, F — UZA 1983.11.16.03 (left), Gl and G2 — UZA 1983.11.19.01, HI and H2 — UZA C2/86, II and 12 — UZA C3/86, Jl and J2 — UZA C4/86, Kl and K2 — UZA C5/86. © Biodiversity Heritage Library, http://www.biodiversitylibrary.org/; www.zoologicalbulletin.de; www.biologiezentrum.at

Indigenous hares of Italy 85

Fig. 12: Camera lucida drawings of occlusal sections of P/3 of different adult specimens of L. corsicanus. Al and A2 — BM 19.7.7.2341, Bl and B2 — BM 98.2.9.1, CI and C2 — BM 78.7.3.4, D — BM 8.9.30.1 (right), El and E2 — MZ 10870, Fl and F2 — MZ 10871, Gl and G2 — MZ 11526, HI and H2 — MZ 11588, II and 12 — INBS 3830, Jl and J2 — AMNH 160956, Kl and K2 — AMNH 160959. © Biodiversity Heritage Library, http://www.biodiversitylibrary.org/; www.zoologicalbulletin.de; www.biologiezentrum.at

86 F. Palacios

This distance is near that existing between L. europaeus—L. castroviejoi (45.31) and

L. castroviejoi— L . granatensis (42.74). The least differentiated species according to the selected variables are the pairs L. corsicanus— L. granatensis (34.98), L. europaeus—L. granatensis (29.93), and L. castroviejoi— L. corsicanus (14.97). Thus, this last pair shows the biggest phenetic similarity according to the analysis. The af- finity between L. corsicanus and L. castroviejoi is supported by other kinds of information such as skull morphology and pelage color and pattern. However, they show some important morphologic differences, particularly as regards the enamel structure of P/3 where L. castroviejoi shows unique characters within the context of the genus Lepus and L. corsicanus resembles the other European species. Fig. 11 and 12 show different P/3 drawings of L. castroviejoi and L. corsicanus corresponding to adult specimens in which the lack or extreme smallness of the anteroflexid in L. castroviejoi is noteworthy in comparison with the regular Lepus anteroflexids of L. corsicanus. It is also noteworthy that in P2/ L. castroviejoi usually has a very slight mesoflexus while L. corsicanus normally has a well-defined mesoflexus as in

Fig. 5, a.

Discussion

All the comparative analyses made in this paper provide clear evidence of the marked phenetic differentation between L. europaeus and L. corsicanus in Italy. A number of discriminant variables and characters exist which, without exception, separate all the specimens of each species. The maximum distinctiveness between the two species was found comparing the skull size, cranial characters, and pelage color and pattern. Therefore, the indigenous hares of Italy previously identified as L. europaeus are here considered to be two different species (L. europaeus and L. corsicanus).

Each species is very homogeneous morphologically. The phenetic characteristics of all the specimens of L. corsicanus from Corsica, Elba Island, the southern Italian Peninsula, and Sicily fully correspond to those of the holotype from Corsica which belongs to an introduced population. In the case of L. europaeus there is no phenetic difference between the old specimens studied in this paper and specimens recently collected in Italy, originating mostly from restocking with hares from other Euro- pean countries. This finding is also supported by comparison with some brown hare specimens indigenous to other countries around Italy (Switzerland, Croatia, France, Spain) studied in different collections. Phenetic distinction within other pairs of hare species from SW Europe has been based on similar characters (Palacios 1989). In the case of L. europaeus and L. granatensis these characters are external measurements, skull size and pelage pattern, and in the case of L. europaeus and L. castroviejoi they are mainly pelage pattern and dental morphology. In these cases geographic patterns of distribution were also parapatric (Palacios & Meijide 1979), except for the pair L. europaeus and L. castroviejoi which, having a basically parapatric pattern may also show a reduced local sympatry (Bonhomme et al. 1986). Natural geographic hare distribution patterns in Europe are now altered. This situation also affects Spain where, over the last 20 years, artificial introductions of L. granatensis have been made in the area of L. europaeus. © Biodiversity Heritage Library, http://www.biodiversitylibrary.org/; www.zoologicalbulletin.de; www.biologiezentrum.at

Indigenous hares of Italy 87

Biochemical characters have also been very useful to distinguish among Spanish hares (Palacios 1979, Bonhomme et al. 1986) but this kind of results is not yet available for the comparison of Italian hares. As regards the taxonomic relationships among SW European hares resulting from the stepwise discriminant analysis based on skull measurements, the four species appear well-differentiated. L. castroviejoi and L. corsicanus are the closest but they differ significantly in some skull variables such as DFZ, TL, ENL, IFL and TBL which are smaller in L. corsicanus. Moreover, there are also differences in pelage color and in dental characters which are considered relevant to lagomorph phylogeny, particularly those related to the anteroflexid in P/3. On the basis of these differences

I consider that it is most appropriate to maintain the species validity of L. corsicanus and L. castroviejoi pending further morphological, biochemical and molecular work.

The enamel pattern in L. corsicanus dentition in Italy is uniform, for instance the anteroflexid in P/3 is large in all localities, including those from Corsica, Elba

Island, and Sicily and is never slight or absent as it is in L. castroviejoi. So it is difficult to support the view that L. corsicanus is the result of an old introduction of L. castroviejoi specimens, or vice versa, because the dental states present in this species do not exist in L. corsicanus. Data from Sicily do not support the idea of hare introduction by the Greeks in the fifth century BC in the terms mentioned by Corbet (1986) because there are no L. europaeus specimens from Sicily in collections. If the

Greeks had carried out any artifical introduction in Sicily, it could have been with L. corsicanus specimens captured on the southern Italian mainland. However, in the current state of knowledge it is difficult to make any hypotheses regarding the presence of L. corsicanus in Sicily because it could have reached the island by natural means.

The present areas of L. castroviejoi and L. corsicanus seem residual. On the base of their phenetic similarity it could be hypothesized that these species probably had a common ancestor occupying a large distribution area in SW Europe before the expansion of L. europaeus. The paleontological material should be studied again in the light of the characters currently available permitting the species distinction, particularly those related to tooth morphology, in order to provide new light on the zoogeography and evolution of this group.

L. corsicanus is a taxon which requires revalidation as a valid species and must be added to the catalog of living mammals of the Italian Peninsula. Old World hare taxonomy is currently very confused and L. corsicanus was not the only example of this situation. Therefore, a great research effort to clarify the systematics of this group should be made, considering that species like L. corsicanus may be disguised locally and need identifying because they may be in danger of extinction.

I recommend surveying the L. corsicanus population in Italy and undertaking a species protection plan in the areas where this hare is still living because it seems likely that this species could be threatened at present due to intense hunting and restocking with brown hares throughout the whole of its distribution area. I hope that these comments stimulate sufficient interest in Italy and France to ensure that conservation measures and actions to save this interesting hare will be undertaken soon. © Biodiversity Heritage Library, http://www.biodiversitylibrary.org/; www.zoologicalbulletin.de; www.biologiezentrum.at

88 F. Palacios

Acknowledgements

I would like to express my gratitude to Dr Maria Luisa Azzaroli (Museo Zoológico de La Specola, Firenze), Dr Michela Podestá and Dr Luigi Cagnolaro (Museo Cívico di Storia Naturale, Milan), Dr Silvano Toso (Istituto Nazionale di Biología della Selvaggina "Alessan- dro Ghigi", Bologna), Giuliano Doria (Museo Civico di Storia Naturale Giacomo Doria, Genova), Dr Gordon B. Corbet, Dr Ian R. Bishop, Paula Jenkins, Jean Ingles, and Daphne Hills (The Natural History Museum, London), Dr Francis Petter and Dr Michel Tränier (Muséum National d'Histoire Naturelle, Paris), Dr Michael Carleton (National Museum of Natural History, Washington), Dr Guy G. Musser and Wolfgang K.-H. Fuchs (American Museum of Natural History, New York) for allowing me to study specimens in the collections in their charge. I thank Dr Sebastiano Salvidio (Istituto di Zoologia della Universita, Genova) and my Spanish colleagues Solis Fernández, Julio Gisbert, Gerado G. Tapia, Jorge F. Orueta, Dr Santiago Reig, Dr Paulino Fandos, Dr Annie Machordon and Laura Barrios for their help during the study of specimens and statistical work. Finally, thanks also to the Departments of International Affairs of the CSIC, CNRS, CNR, and The Royal Society for managing and financing travel fellowships to visit museum collections.

Zusammenfassung

Hasen von der italienischen Halbinsel und Sizilien werden üblicherweise einer einzigen Art (Lepus europaeus Pallas, 1778) zugeordnet. Eine taxonomische Untersuchung, basierend auf Größe, Morphologie und Fellmerkmalen autochthoner Exemplare aus Museumssammlungen, ergibt dagegen, daß sie zwei verschiedenen Arten zugehören (L. europaeus und L. corsicanus de Winton, 1898). Die Kartierung der Fundorte beider Hasen zeigt, daß sie natürlicherweise parapatrische Areale einnahmen, L. europaeus im Norden und L. corsicanus im Süden Italiens, mit einer Trennungslinie zwischen Sienna und Rom. Beide Arten sind phänetisch klar getrennt, ihre Unterscheidung ist mit Hilfe univariater und multivariater Statistiken möglich. Die phänetischen Charakteristika jeder Art in ihrem natürlichen Verbreitungsgebiet sind be- merkenswert gleichförmig. Zwischenformen wurden nicht gefunden; Exemplare aus benach- barten Orten weisen keine intermediäre Eigenschaften auf. Bei den Unterscheidungsmerkma- len, die diese beiden Arten voneinander trennen, spielt die Farbe des Rückenunterfells adulter Exemplare eine besondere Rolle, die bei L. europaeus immer weiß und bei L. corsicanus grau ist. Die Schädelgrößen unterscheiden sich in einigen Abmessungen, wobei L. corsicanus kleiner ist als L. europaeus. Ein wichtiges Zahnmerkmal ist die posteriore Kontur im Quer- schnitt des ersten oberen Schneidezahns, der bei L. europaeus konkav, bei L. corsicanus konvex ist. Statistische Vergleiche der Schädel- und Zahnmerkmale weisen hohe Chiquadrat- werte und sehr geringe Ähnlichkeiten auf, was auf einen deutlichen Unterschied zwischen beiden Arten schließen läßt. L. corsicanus zeigt im Vergleich mit anderen SW-europäischen Hasen die größte phänetische Ähnlichkeit mit L. castroviejoi und die geringste mit L. europaeus. Die Ergebnisse lassen vermuten, daß L. corsicanus und L. castroviejoi einen gemein- samen Vorfahren hatten, der vor der Ausbreitung von L. europaeus ein großes Verbreitungs- gebiet in SW-Europa zwischen Italien und Spanien einnahm.

References

Angermann, R. (1983): The taxonomy of Old World Lepus. — Acta Zool. Fennica 174: 17-21.

Bonhomme, F., J. Fernández, F. Palacios, J. Catalan & A. Machordon (1986): Ca- ractérisation biochimique du complexe d'espéces du genre Lepus en Espagne. — Mam- malia 50: 495-506. Cabon-Raczynska, K. (1964): Studies on the European Hare. III. Morphological Varia- tion of the Skull. — Acta Theriol. 9: 249-285. Corbet, G. B. (1986): Relationships and origins of the European lagomorphs. — Mammal Rev. 16 (3/4): 105-110. Ellerman, J. R. & T. C. Morrison-Scott (1951): Checklist of Palaearctic and Indian Mammals (1758 to 1946). — Trust, of the Brit. Mus. (Nat. Hist.) London, 810 pp. © Biodiversity Heritage Library, http://www.biodiversitylibrary.org/; www.zoologicalbulletin.de; www.biologiezentrum.at

Indigenous hares of Italy 89

Flux, J. E. C. (1983): Introduction to taxonomic problems in hares. — Acta Zool. Fennica 174: 7-10.

Flux, J. E. C. & R. Angermann (1990): The Hares and Jackrabbits. — In: Chapman, J. A. & J. E. C. Flux: Rabbits, Hares and Pikas, Status survey and Conservation Action Plan, 61-94. IUCN, Gland, Switzerland. Hoffmann, R. S. (1993): Order Lagomorpha. — In: Wilson, D. E. & D. M. Reeder: Mam- mal Species of the World, A taxonomic and geographic reference, 807—827. Smithsonian Institution Press, Washington and London. Miller, G. S. (1912): Catalogue of the Mammals of Western Europe. — British Museum, London, 1019 pp. Palacios, F. (1979): Análisis cromosómico, carga de DNA y electroforesis de las proteínas de las liebres españolas. — Doñana, Acta Vert. 6: 203—215. Palacios, F. (1989): Biometrie and morphologic features of the species of the genus Lepus in Spain. — Mammalia 73 (2): 227—264. Palacios, F. &M. Meijide (1979): Distribución geográfica y hábitat de las liebres de la Península Ibérica. — Nat. Hisp. 19: 1—40.

Palacios, F., J. F. Or ueta &G. G. Tapia (1989): Taxonomic review of the Lepus europaeus

group in Italy and Corsica. — Abstract of papers and posters, V ITC, Rome, 1: 189—190. Petter, F. (1961): Eléments d'une revision des Liévres européens et asiatiques du sous-genre Lepus. — Z. Säugetierk. 26: 30—40.

Ridgway , R. (1912): Color standards and color nomenclature. Washington D. C, 43 pp, LIII Pi. Toschi, A. (1965): Fauna dTtalia VII, Mammalia (Lagomorpha-Rodentia-Carnivora-Artio- dactyla-Cetacea). Edizioni Calderini Bologna, 467 pp. Vigne, J. D. (1988): Les Mammiféres post-glaciaires de Corse: étude archeozoologique. 26e suppl. á Gallia Préhistoire, CNRS éd., 330 pp. Winton, W. E. de (1898): On the Hares of Western Europe and North Africa. — Ann. Mag.

Nat. Hist. London, 1: 149-158.

Dr. Fernando Palacios, Museo Nacional de Ciencias Naturales (CSIC), José Gutiér- rez, Abascal 2, 28006 Madrid, Spain. E-mail .

Appendix 1: L. corsicanus material examined

Istituto Nazionale di Biología della Selvaggina, Bologna (INBS) 3306 Mongiana, Catanzaro, Italy, 18 Apr 1975, mounted skin, skull inside, female, juvenile.

3830 Mongiana, Catanzaro, Italy, Oct 1974, skin and skull, age III.

Museo Zoológico de "La Specola", Firenze (MF) 10870 Saponara, Milazzo, Sicily, 3 Dec 1883, mounted skin, skull, male, age IV. 10871 Vicari, Palermo, Sicily, 12 Dec 1883, mounted skin, skull, female, age IV. 11526 Aleria, Corsica, 3 Oct 1889, mounted skin, skull, male, age IV 11588 Isola d'Elba, Italy, 25 Feb 1877, mounted skin, skull, male, age IV 11590 Isola d'Elba, Italy, 12 Feb 1877, mounted skin, skull inside, female, adult. 11592 Foggia, Italy, 21 Feb 1908, mounted skin, skull inside, male, adult.

Museum National d'Histoire Naturelle, Paris (MNHN) 1894-37 Corsica, mounted skin, adult. 1919-670 Bastia, Corsica, mounted skin, adult.

1962-2546 Corsica, 29 Sept 1910, skull, age II. © Biodiversity Heritage Library, http://www.biodiversitylibrary.org/; www.zoologicalbulletin.de; www.biologiezentrum.at

90 F. Palacios

Natural History Museum, London (BM)

8.9.30.1 Surroundings of Rome, Italy, 13 Jan 1900, skin and skull, male, age IV. 19.7.7.2341 Corsica, Nov 1883, skull, age IV. 78.7.3.4 Bastia, Corsica, Jan 1875, skin and skull, male, age IV, Holotype. 98.2.9.1 Marsala, Sicily, Jan 1898, skin and skull, male, age IV.

98.2.9.2 Marsala, Sicily, Jan 1898, skin and skull, age III.

American Museum of Natural History (AMNH) 160956 Abruzzo, Lago Fucino, Italy, Dec 1935, skull, female, age IV. 160959 Monti del Gargano, Vieste, Puglia, Foggia, Italy, Dec 1933, skull, female, age IV.

Appendix 2: L. europaeus material examined

Museo Zoológico "La Specola", Firenze (MF) 10871 (dupl) This skull has been extracted from the mounted L. corsicanus skin n. 10871 and belongs to L. europaeus. As the true skull of the specimen n. 10871 was already prepared in collection,

I have included the L. europaeus skull in the study with the number 10871 (dupl), assuming

that it could have a similar date to 10871 (around 1883) and could be a local specimen (pro-

bably of the surroundings of Florence). It seems likely that the L. europaeus skull was used to facilitate the mounting of the specimen n. 10871. The skull n. 10871 (dupl) corresponds to relative age IV. 10872 Alpe della Luna, Arezzo, Italy, 9 Nov 1885. This specimen consists of a mounted skin of L.

europaeus with skull inside, and another skull which is prepared in collection and also belongs to L. europaeus. As both the mounted skin and the prepared skull have the same label

data, it is likely that the mounted skin and the skull inside do not belong to the same speci-

men, it being possible that this second skull was used for naturalization only. Male specimen. Both the mounted skin and the prepared skull correspond to age IV.

10873 S. Casciano di Bagni, Siena, Italy, 10 Jan 1900, skull only, male, age III. 10874 Rassina, Arezzo, Italy, 31 Jan 1890, mounted skin, prepared skull, male, age IV.

11523 Maiano, Firenze, Italy, 18 Sept 1899, skull only, male, age III.

11524 S. Casciano di Bagni, Siena, Italy, 10 Jan 1900, skull only, female, age III. 11525 Gabbiano, Greve, Firenze, Italy, 23 Dec 1899, skull only, male, age IV. 11527 Surroundings of Trieste, Italy, Jan 1889, skull only, age IV. 11584 Alpi di Domodossola, Italy, Sept 1876, mounted skin, skull inside, male, juvenile. 11585 Camugliano, Toscana, Italy, 2 Nov 1880, mounted skin, skull inside, male, adult. 11586 Torino, Italy, 10 Dec 1883, mounted skin, skull inside, male, adult.

11587 Greve, Firenze, Italy, 20 Aug 1900, mounted skin. It is written on the label that this specimen

has a detached skull but it is not in collection. Examining the skin it seems that the skull

is not inside either. Female, juvenile.

11591 S. Maria a Monte, Bientina, Italy, 20 Sept 1883, mounted skin, skull inside, female, adult. 11594 Firenze, Italy, Jan 1873, mounted skin, skull inside, female, adult. 11595 Firenze, Toscana, Italy, Oct 1838, mounted skin, skull inside, male, adult. 11596 Firenze, Toscana, Italy, Dec 1837, mounted skin, skull inside, male, adult. 11597 Domodossola, Italy, June 1887, skull only, age III.

11598 Domodossola, Italy, June 1887, skull only, age III.

11599 Domodossola, Italy, June 1887, skull only, age II.

Natural History Museum London (BM) 19.7.7.2486 Genova, Italy, Feb 1884, skull only, age IV. 98.10.2.18 Surroundings of Siena, Italy, 2 Feb 1898, skin and skull, male, age IV. 98.10.2.19 Siena, Italy, 13 March 1898, skin and skull, male, age IV.

National Museum of Natural History (US NM) 153399 Siena, Italy, 29 Dec 1898, skin and skull, female, age III. 153400 Siena, Italy, 15 Nov 1898, skin and skull, female, age IV. © Biodiversity Heritage Library, http://www.biodiversitylibrary.org/; www.zoologicalbulletin.de; www.biologiezentrum.at

Indigenous hares of Italy 91

Appendix 3: Alphabetical list of abbreviations of body and skull measurements a) Body E Ear length HB Head and body length HF Hindfoot length T Tail length b) Skull ANW Anterior nasal width, taken between the prominences of the nasal processes of the incisive bones. DPFP Diameter of the posterior foramen palatinum. DFZ Diameter of foramen zygomaticus.

T"\TT* C1 TA /Í DIRMM Distance between the incisor root and sutura incisivomaxillaris. ENL External nasal length FIL Foramen incisivum length.

T?„ ' • • '1,1 FIW roramina incisiva width. FTL Facial tubercle length. FUISL First upper incisor section length. FUISW First upper incisor section width. HPMM Height of processus muscularis mandibulae, taken between the anterior margin of processus articularis and the tip of the process, perpendicular to the margin. INL Internal nasal length. LCTRL Lower cheek-tooth row length. MH Mandible height.

' X /f I'll I A, 1 J. I C _ , T „ , 1 MLCP Mandible length, taken from the capitulum. NPB Narrowing of the palatine bridge. PL Palatal length. PNW Posterior nasal width. PPW Postpalatal width. PZW Posterior zygomatic width. Radius of the first upper incisor curvature. RW Rostral width. SFW Smallest frontal width. TBL Tympanic bulla length. TBW Tympanic bulla width. TL Total length. UCTRL Upper cheek-tooth row length. WFT Width between facial tubercles. WSP Width between supraorbital processes.